In this project the mechanisms of mutagenesis are investigated through a detailed study of its specificity. DNA sequence information is gathered on all the classes of mutations that occur: base substitutions, frameshifts, deletions, duplications, insertion elements, complex rearrangements, etc. These classes have their own dependence on the local DNA sequence and generally result from different mutational pathways. The specificity of mutation thus provides a way to analyze and separate the various components of mutation. We use the lacI gene of the bacterium E. coli as a mutational target. The gene codes for the repressor of the lac operon and forward mutations to lacI are scored based on their constitutive expression of the operon. The lacI genes (typically several hundreds at a time) are transferred by in-vivo recombination to a single-stranded (recombinant) phage vector and sequenced, producing the mutational spectrum of interest. Comparing spectra in strains affected in various DNA repair/replication pathways is a next important step. In case of defined enzymatic pathways, the spectra provide a direct correlation between mutational classes and their responsible pathways. In case of unknown pathways, the mutational specificity may provide new insights into the affected pathway. So far, we have determined the specificity of mutation in mutH, mutL, mutS, mutT and mutD and wild-type strains of E. coli and have gained insights into the specific contributions of DNA damage, DNA mismatch repair and exonucleolytic proofreading to mutation. In case of induced mutagenesis, the specificity of mutation is a tool to identify both the nature of the premutagenic lesions and the mechanisms by which these lesions are converted into mutations. An example of this is the specificity of mutagenesis by the chemical carcinogen N-acetoxyacetylamino-fluorene, which has enabled us to formulate a molecular model delineating how a single lesion may create different mutations depending on the local DNA sequence.